190 THE NEMERTINI 



communication between the two systems, as Oudemann believed. 

 There is never any anastomosis between the branches, nor any 

 communication between the right and left organ. 



The restriction of the system to a limited region of the body, 

 and the absence of a network, such as occurs in Platyhelmia, is 

 no doubt connected with the existence in the Nemertines of a 

 " vascular or lymphatic system," which brings to the excretory 

 system the material for its activity. There can be no doubt but 

 that the system, though differing in details and in plan from that 

 of Platyhelmia, is descended from it, and belongs to the same 

 category as the " head kidney " of Annelids, which it resembles in 

 its limitation to the stomodseal region of the gut. 



The reproductive organs contrast as much as is possible with 

 those of the Platyhelmia, for in the Nemertines there is nothing 

 resembling the copulatory organs of the former phylum; there are 

 no glands set aside for the formation of egg-cases, no differentia- 

 tion of the ovary into " germarium " and " vitellarium." We have 

 to do with mere sacs (Fig. XVIII. y) containing the products of 

 the proliferation and modification of the epithelial cells which line 

 these sacs, and in due time, when these products are ripe, these 

 sacs push their way outwards through the body wall to form genital 

 ducts, which will ultimately open to the exterior. In a few 

 Metanemertini ovaries and testes occur together and ripen simul- 

 taneously (Prosadenoporus, two sp., Tetrastemma, two sp., Geonemertes 

 spp., and Prosorochmus), while " Borlasia " Kefersteini, Stichostemma 

 Eilhardi are protandric hermaphrodites. 



The genital sacs are coextensive with the midgut, and as a rule 

 are repeated in a regular series, one sac between every two suc- 

 cessive enteric pouches. This regularity is, however, concealed in 

 Amphiporus, in that only some of them ripen at a time, so that in 

 A. pulcher there are only five, at irregular distances. In Carinella 

 and Malacobdella, in which the intestinal pouches are not found, the 

 genital sacs are closely packed together. The pores, as a rule, 

 form a simple linear series above the lateral nerve, but in the two 

 genera just named, and in G. australiensis they form a broad band 

 extending nearly to the mid-dorsal line. In Drepanophorus, owing 

 probably to the great development of the dorsal organs of the 

 body, e.g. the rhynchocoel, the genital organs, like the lateral 

 vessels, seem to be ventral (Fig. XXI.). 



The genital sacs arise either (a) simultaneously with the de- 

 veloping genital cells (as Carinella, Malacobdella, Prosadenoporus, and 

 others) from a group of parenchymal cells which gradually become 

 differentiated into a central mass of " germ cells " and a peripheral 

 membrane of flat wall cells ; or (b) the sacs develop first, and then 

 from some of the epithelial cells the germ cells arise (as in Cere- 

 bratulus, Drepanophorus) by the accumulation of yolk spherules 



